When it comes to the delicate nature of SSB signals, just a "guess" is not the best way to go.
I'm sure people on here have seen this...
[ATTACH=full]34414[/ATTACH]
Even this...
[ATTACH=full]34408[/ATTACH]
Best to use resistor dividers as voltage dividers - to float DC values of Gate voltages to a threshold level to keep the PART off but allow it to turn on without cutoff issues - and try to keep DC - just plain - DC and Keep RF out of them...
NEWLY REVISED
[ATTACH=full]34420[/ATTACH]
Just don't use so much resistance from the 8 volt sources that the SSB words are "Cutoff" - meaning you aren't in the Gates' Sweet Spot to operate as almost linearly - just off - not totally on. Cutoff is caused by lack of voltage to push the MOSFET on - because only part of the words are getting amplified - you need to increase the bias voltage to a higher level. To do that - lower the FEEDER (your resistor from the +8volt source) resistance to a lower range.
- Above: Note the RESISTOR/BEAD RF Isolation section that is for keeping RF out and allowing some linear bleed of current and power into thru and back out of the bias circuit from the RF input going to the GATE. A simple 50 to 100 ohm resistor of 1/4W value carbon Composition can work as your buffer resistor. You just don't want RF leaking into the bias circuit more that it needs to be - so use the RESISTOR/BEAD combination as much as possible to keep RF where it belongs - at the Gate!
BE CAREFUL with this method. Because you can experience DC switch failure, you see, these things will "latch on" and blow even before you even know it's on.
Yes, they do require current to operate - but far less in values than Bipolars. They need voltage to form a field in the Gate to make the device turn on...that requires some current - but not much.
Too much current can cause latching as well as cutoff because there is simply too much power flowing THRU the circuit - so keep the resistance high, to keep the power flow from WASHING AWAY or SWAMPING the RF so it can't get to the Gate...
So, the Graphics above are ARBITRARY - meaning they are not cast in stone...you can use other values like even what the UNIDEN 687 uses in the 520 series they use and have - only it's 8.2K and 4.7K - but that gives you only so much...
Ok to complete this because of time and effort and family life juggling together...
The Self-bias parts - like shown above - the EN1230 / EN369 variants are up to you - they most are for FM or AM work...
But when you want more reliable biasing, you'll have to use sourcing voltages from the on-board regulation system the radio has whether it'd be for the PLL system or the standard 8+volting system - either way, you should bias accordingly but use trimmers too. Not all MOSFET are going to respond in like fashion...
You can find more out about that here - Palomar ERF-2030+ at this link - it's just that time is an enemy of me too, and it sometimes gets in the way. So again I add, that to apply the "divider" is a better option, just not the best for handling variations of signal, voltage, power and other needs of the MOSFET's you decide to drop in there to replace that old worn out Couch of a Final called '2078, or the Refrigerator like the '2312 or the Off-White Kelvinatror - the '1969.
So variations will occur, your mileage may vary, and so therefore add in some trimmers whether on the source side of the feeder line for your voltages, or incorporate the variable in your bias design in the shunt or ground side of the divider - just use a resistor in series with it too to lessen the likelihood of having a bigger failure cause by excessive drive blowing up your variable pot - to pot. So if you blow the Gate, you've applied too much either power or RF drive - more than likely it was from setting idle bias too high and the thing latched on due to the Gate telling the part to turn on. Even when there is no signal. So the Series approach to this helps in getting the Bias current to voltage ratio set properly and low enough to keep the part from latching on by accident or strong RF proximity you didn't plan for.
